Abstract Evidence over the past 30 years has revealed that 70-80% of individuals with schizophrenia exhibit marked neurocognitive deficits on measures of attention, learning and memory, problem-solving, language and sensory-motor skill (e.g., Reichenberg & Harvey, 2007). Particular significance has been attached to these deficits as their severity has been linked to impaired community function, social problem-solving and progress in psychosocial rehabilitation programs (Green et al., 2000; Green et al., 2015). Cognitive remediation (CR) is a type of behavioral intervention that addresses cognitive deficits in schizophrenia by restoring lost cognitive skills or providing strategies for bypassing deficits through task practice. Meta-analyses (Wykes et al., 2011; McGurk et al., 2007) have revealed that cognitive remediation is a validated approach to improving cognitive function in schizophrenia, however a lack of precision regarding the active elements of the intervention have prevented its recommendation as a standard treatment for the illness (Dixon et al., 2010). The present three-year proposal seeks to identify cognitive training mechanisms that are most effective at improving cognitive function in schizophrenia by comparing two different systematic programs of CR with different foci: drill-and-practice exercises vs. compensatory strategies. Both programs have strong preliminary empirical support. One-hundred and thirty-five clients diagnosed with schizophrenia or schizoaffective disorder will be randomly assigned to one of three groups: a neuroplasticity-based, drill-and-practice program of computer-assisted cognitive training exercises designed to restore lost cognitive capacity; a manualized strategy training method for bypassing deficits in cognition, or a computer games control condition. Study measures, organized according to an experimental therapeutics approach, with targets distinguished from outcomes, will assess generalization of any observed training effects.